Dot gain compensation\profiles

[destryer]

We run newsprint with muddy dot gain peaking at 85%. I have created a (simulation) dot gain compensation curve to remove the gain in the 60-90% range. I need to build new profiles that show this curve applied so everyone will see the changes and proofs will match the press. In a normal workflow you would run your target uncalibrated to build your profiles would this be a correct way to correct for gain and build the profiles?

 

[gordo]

We run newsprint with muddy dot gain peaking at 85%. I have created a (simulation) dot gain compensation curve to remove the gain in the 60-90% range. I need to build new profiles that show this curve applied so everyone will see the changes and proofs will match the press. In a normal workflow you would run your target uncalibrated to build your profiles would this be a correct way to correct for gain and build the profiles?

Looking forward to the answers you get.
I would just use a tone compensation curve applied to the plate to normalize "dot gain" in the presswork, then characterize the presswork and build a profile to drive the proofer so that it better simulates the presswork. Some newspaper printers just run their proofs on butt roles this way.

best, gordon p

my print blog here: Quality In Print

 

[toronar]

Hm ... this is the way I create profiles for our presses:

1. make sure the plates are linear, i.e. 10% comes out as 10%, 50% as 50% and so on
2. print a testchart from these linear plates (the testchart contains several color strips, an ECI 2002 chart and test strips for every color from 0% to 100%) and achieve the best Lab-values for your primary colors
3. measure the dot gain in the test strips, compensate accordingly to a standard
4. apply the compensation and make plates again, print again to achieve the same Lab-values of your primary colors
5. measure test strips for dot gain, if satisfactory go to 6 else repeat 3-5
6. measure ECI 2002 chart and create a profile from it - for the proof and as an ICC/Devicelink-profile to convert pictures etc. with

Repeat for every type of paper (coated, uncoated, yellowish, ...) and every ink, every press ... and if you have too much time: for summer/winter, different blankets, every paper you use, different color setups ...

Don't forget to re-linearize your plates once in a while (we do it around once a week and if we have a different batch of plates).

That should be it - the very short version

 

[gordo]

There's no need to linearize the plate first in order to create dot gain compensation curves for the plates. It is only necessary that the plates are consistently imaged. You can measure the response from your uncalibrated plates and build a curve to bring the presswork to align with your target (i.e. SNAP in this case).
Since this is for newspaper there's no need, in this case, to repeat for every type of paper. There is also no need to re-linearize the plates once in a while, only to make sure that they are imaged with the correct curve and that the tone values on plate are correct ( e.g. a 50% request results in a 42% tone on plate).
I think that a dedicated press run should be done to build your curves and get the data to build your profiles, however, in newspaper work, a good strategy that may also work is to run test targets in the form of a small, or several small advertisements. This gets you the real estate in the pressrun without doing a dedicated run.

best, gordon p

my print blog here: http://qualityinprint.blogspot.com/

 

[toronar]

I guess you are right gordo, if it is only for one sort of paper.
I work in the offset printing business, and I find it less cumbersome to start off with linear plates if you have to profile/compensate for more than one paper/press. You can do it without the linearization first, of course.

Our problem lies within "consistently imaged".
Depending on the batch of plates from the dealer ... our values may change upto 5% on the plates. According to Fuji (CTP imagesetter, plates and finisher) this is normal. *sigh*
The change sometimes only occurs <20% (or just >60%) while the rest stays fine. Re-linearization is by far faster in this situation with profiles for 3 presses and around 7 different printing conditions per machine
And of course re-linearization is only done when the target values do no longer match - we output a special mark on the plates just for checking that.

Either way works fine, though.

 

[otherthoughts]

I guess you are right gordo, if it is only for one sort of paper.
I work in the offset printing business, and I find it less cumbersome to start off with linear plates if you have to profile/compensate for more than one paper/press. You can do it without the linearization first, of course.

Web offset or Sheet-fed offset?

Regards
otherthoughts

 

[mklafke]

I think there is a need to calibrate the plate if you want to work like the ISO-Norm ISO 12647-2 dicdates.
In the Media Standard Print is a entry File value Data = File Value Plate.

Its easier to correct the Curve of the Plate than the Calibration Curve.

If you´re using different Plates or if you have 2 or more Platesetters with crossover Workflow its much more comfortable to work with PlateCurves.

 

[meddington]

I think there is a need to calibrate the plate if you want to work like the ISO-Norm ISO 12647-2 dicdates.

Not sure if you're implying that calibrating the plate is a requirement of ISO 12647-2? The standard itself doesn't mandate this.

In the Media Standard Print is a entry File value Data = File Value Plate.

This wording does not necessarily mandate plate calibration, and is a little ambiguous File value data could imply either before or after tone curves (assuming the latter). Either way, consistency is more important than the raw plate value.

Its easier to correct the Curve of the Plate than the Calibration Curve.

If you´re using different Plates or if you have 2 or more Platesetters with crossover Workflow its much more comfortable to work with PlateCurves..

The terminology here is a bit confusing here. "Plate Curves" and "Calibration curves" could mean the same thing to some people. I'm infering "calibration curve" to refer to a linearization procedure and "plate curves" as a tone adjustment to meet the desired target print conditions. Is this your intended meaning?

 

[destryer]

Linearization

Linearization

We also linearize our plates since we have 2 plate setters and need them to match. Not including linearization in our plates means losing another place to control your workflow. I've also had issues with plate materials and linearizing the plates can quickly compensate for supply quality fluctuations without reprofiling.

 

[glenncolorguy]

Dot Gain Compensation_Profiles

Dot Gain Compensation_Profiles

This is the correct approach, at least to a point.

Profiles for proofing and simulation should reflect actual printing conditions, so if you are using a negative curve in your platemaking process to reduce dot gain on press, this should be reflected in your proofing profile, most likely by measuring a profiling chart printed under these conditions.

However, this approach puts too much focus on a single press run, which may not be typical, and does nothing to get you closer to an industry standard.

The best approach is to proof using an industry standard (eg. IFRA or ISO Newsprint) profile for proofing, and adjusting your printing process to match this result.

 

[Mike Strickler]

I just typed a reply that vanished--thought I was signed in but apparently not--and now read that Glenn has written the same comment almost word for word. Let me add that your plate reader is probably responsible for a lot of the "fluctuation" you see, notwithstanding the mfr's discliamer. Always try to root out the the problem by starting linear and measuring (on paper) the actual dot gain gain on the three-color grays at three or more points (e.g., 25, 50 75). If your press and substrate are reasonably "standard" just use a good off-the-shelf newsprint profile (e.g. SNAP); this can be edited further to reflect additional gain your press may give, but be careful to keep the edits simple, no more than a gentle tug on the global gradient curve at around 50%. If you're not sure about this or have no experience just stick with the standard profile and a simple transfer curve at the platesetter.

 

[martiald]

Hi. This is the techniques i used for gravure, litho, flexo, and now dry-offset...
Basically we have to think backwards.
We need to be able to anticipate the print result from the computer file (as well, more and more approvals are done through PDF)
1/Linearization: You need to ensure that 10% in the file is 10% on the plate. At this stage, it doesn't matter what it is on the print.
2/Once the linearization is done, and you trust your plate, print it. Measure your dot gain each 5% step.
3/compensate: from those readings, create a compensation curve through the RIP (not the soft). Eventually, the 10% in the file will become 6 ot 7 on the plate, but then will gain 4 to 5% on the print.

4/To push it further is the colour profiling. We have to bear in mind that the dot gain compensation, or gradation, is not related to the colour, but simply to your dot size. Each screen and amgle will require a different dot gain curve, but often you can work from 1 standard. Colour profiling is completemy different, it measures the colour interactions and transparencies. For this the measurement needs to go through a spectrophotometer, not a basic densitometer. but this is another story.

I hope this helps a bit

rgds.

 

[pam_add]

I am confuse please little explain
I must the linerize the press like 50% print 50% on paper or i must use iso dot gain ref like 50% in file must print 64 if yes how must i correct my images in photoshop since on my eizo i see diffrence between my calibrated screen and the pantone procees chart which printed using the iso dot gain reference